Automate post-task reflection and memory updates in JaRVIS journal files.
npx skills add https://github.com/epicrunze/JaRVIS --skill jarvis-reflect
Please help me install this Agent Skill. Skill: jarvis-reflect Source: https://github.com/epicrunze/JaRVIS/tree/main/skills/jarvis-reflect Command: npx skills add https://github.com/epicrunze/JaRVIS --skill jarvis-reflect
After completing a meaningful task, you need a disciplined way to capture outcomes, lessons, and decisions so your AI agent improves over time and its identity evolves accurately.
Run /jarvis-reflect after completing a meaningful task to write a structured entry and persist memory updates.
High-intent search queries and answers about installing and using this skill.
Post-task reflection and memory sculpting solves this by capturing outcomes, lessons, and decisions after meaningful work. It writes a structured journal entry and updates preferences to guide future AI behavior accurately.
Memory sculpting works by updating preferences and decisions with rationale to guide future behavior. It persists these updates within memory files in the data directory, ensuring the AI agent evolves its identity accurately across tasks.
You must run the reflection command to write a frontmatter-annotated journal entry, update preferences and decisions, validate the data, and commit changes. This structured workflow ensures memory updates persist accurately in the data directory.
Yes, identity-aware evolution triggers identity updates and tracks milestones when applicable. Running the reflection command after meaningful tasks updates memory and journal files to ensure your AI agent's identity evolves accurately over time.
The best way to validate memory updates is to run the validation command before committing changes in the data directory. This ensures the frontmatter-annotated journal entry and updated preferences are correctly formatted for memory persistence.
epicrunzeValidate JaRVIS data directory journals, memories, identity fields, and growth logs.
felix-howardGenerate journal entries documenting code changes and session reflections in ./docs/journals/.
dlantsIndex local Markdown files and retrieve semantic search results with contextual information.
krishagelStore and retrieve user preferences with confidence scores in structured JSON files.
